WO2018091124A1

WO2018091124A1 - Method for transmitting information

Info

Publication number: WO2018091124A1
Application number: PCT/EP2017/001243
Authority: WO
Inventors: Hristo PETKOV; Thomas LAUTENBACHER; Thomas Kauppert; Klaus Gottschalk
Original assignee: Diehl Metering Systems Gmbh
Priority date: 2016-11-16
Filing date: 2017-10-24
Publication date: 2018-05-24
Also published as: DE102016013654A1; CN109891792A; EP3542482A1; US20190253928A1

Abstract

During the repeated transmission of information in the form of a coded package (14) from a transmitter (11) via a wireless connection (13) to a receiver (12), a fault-free receipt package (14') is quickly provided when fault-free partial packages (15) from the package (14) are lined up as partial packages (15) of the fault-free receipt package (14'). For this, partial packages (15), which are allocated to each other on the receiving side, of successively received and temporarily stored packages (14) can be compared with each other and if they match can be incorporated in the receipt package (14') to fill gaps. Or the successive packages (14) are split up on the transmission side to form partial packages (15) to which error identification codes (16) are assigned, which partial packages (15) are included in the receipt package (14') as soon as they have been correctly received in a package (14).

Description

Method for transmitting information

The invention relates to a method according to the preamble of the independent claim.

The information is coded telegrams, hereinafter referred to as packages, for example for radio remote control of building services such as heating, local lighting, movement of blinds and shutters or locking garden and garage doors; but also for the radio transmission of consumption measured values (in particular for gas, water, heat or electricity) from individual measuring points to a common receiver memory (also referred to as concentrator).

Such packets of predetermined length are transmitted identically several times in succession and buffered in a receiver, that is to say in particular in the concentrator, in order to compare the successive packets with one another. If several of them match, these packets are considered correct as a result of undisturbed information transfer. If, however, only one bit deviates from the bit value at the same position in the previously received packet, all packets previously buffered with this measured value are rejected as unsafe; and wait for a new receive sequence from at least two full match control packets to be cached.

On the chronically congested publicly available radio channels in the ISM bands, the probability is low that even a single such buffer cycle leads to a correctly received, ie undisturbed reception packet; and the likelihood decreases with the length of the packets, so the possibility that in the course of radio transmission, the noise environment introduces a bit error in the packet. The need for continuous repeating the radio transmission of a package not only leads to the increase in the duration to the trouble-free reception of a complete package, but also represents a significant burden on the battery in the measuring station transmitter and an additional load on the transmission channel. In recognition of such circumstances, the present invention is based on the technical problem of accelerating the detection of error-free transmission of packets and thereby being able to reduce both the loading of the transmission current source and the channel occupancy.

This object is achieved by the essential features specified in the main claim. Thereafter, the packets in the receive memory are no longer checked for coincidence overall due to correct transmission, but according to the invention each packet is parcel-by-packet.

For this purpose, in the context of the present invention successively received packets can each be divided into subpackets; and from matching, therefore correctly received sub-packets, the undisturbed receive packet is lined up to fill the gap. Or even the package containing the information is broken down into subpackets, which are then each sent with an error detection code; and therefore correctly received subpackets, the undisturbed receive packet is lined up in a gap-filling manner.

The subpackets do not have to have the same length; Rather, the sub-packet lengths may vary statically or dynamically over the length of their packet. It only has to be ensured that at least two packet repetitions have the same partial packet definitions with respect to one another, if on the receiver side the error freedom of precisely this partial packet is to be concluded from consistently repeated partial packets.

When a transmission error occurs so no longer an entire package is discarded, but discarded at most currently incorrectly received sub-packets; from the stringing of correctly received subpackets, the complete, ultimately undisturbed receiving packet is lined up in the concentrator or the like temporary storage.

The subpackets that make up a package are inevitably shorter than the complete package. Thus, the likelihood that individual subpackets will be received incorrectly is much less than with respect to the overall packet length. The undisturbed receive packet composed of error-free subpackets is therefore available more quickly than a packet examined on the receive side for complete freedom from errors. In order to find the error-free received subpackets, as already mentioned, the completely received packets can be successively stored in the receiver and viewed in each case their associated subpackets. Such corresponding subpackets, which now coincide, are considered correctly received; because the probability that in the short sub-packets mutually associated bits have experienced the same disorder is sufficiently low. If then the assembled correct sub-packages again give a complete reception package, this is undisturbed.

It is even faster to carry out the partial packet-wise check according to the invention if the transmission-side subdivision of the packet into subpackets already takes place, and an error detection code is assigned to each of the subpackets. Such a can be a so-called CRC checksum, or even, for example, as practiced in WMBUS, a 3-out-of-6 encoding.

A subpackage may also be equipped with multiple error detection capabilities simultaneously, such as multiple contiguous 3-out-of-6 encodings or multiple CRC checksums. In any case, each of their correctly received sub-packets need to be buffered only from consecutively received packets, until they have finally lined up in a gap-filling manner to form a correct receiving packet.

If, on the transmitting side, subpackets equipped with at least one error detection code are received with errors, they do not have to be discarded. Rather, the two mentioned variants can now be combined with one another in the sense that the faulty subpacket in turn - and also the corresponding subpacket used for comparison at least of the next following packet - are subdivided into sub-subpackets. Of these, only one will be faulty, the others can already be inserted into the still open defect in the receive packet. Such sub-decomposition can be further staggered to exploit useful packages.

Application of such sub-decomposition to successively correctly received sub-packets leads to a desirable improvement in the residual error probability.

In any case, so waiving a conventional assessment of the received packets as a whole, from the several consecutive disturbed received packets whose undisturbed received sub-packets picked out, cached and finally assembled into an undisturbed reception package. It is therefore no longer interesting whether two successive packets do not match and therefore should actually be discarded - instead, from each of these useless packets their usable sub-packages are now picked out, in order to arrange an undisturbed receive packet. This is achieved faster than waiting for a total of undisturbed received packets; whereby the energy-related and channel-specific resources are relieved, because the origin packets must be sent less frequently until a usable receive packet is reached. The advantage of assembling the receive packet according to the invention from undisturbed received subpackets is the greater, the longer the original information message (ie the packet), because the risk of interference with the packet length increases; which is why very long packets in the conventional full assessment have almost no chance to be transmitted over one of the heavily interfered channels.

Mathematically, an improvement in the likelihood of success in terms of a usable available receive packet - and thus the system QoS (compliance of the quality of a communication service in a given noise environment with the requirements of the user) - by at least 20% detectable.

Additional developments of the invention and its alternatives will become apparent from the other claims and, taking into account the advantages thereof, from the following description of preferred embodiments of the invention. In the drawing shows by way of example, limited to symbolic sketches with illustratively unrealistically few retransmissions and partial packets using the example of a unidirectional radio link,

1 shows an information transmission with only receiving side partial packet formations on the repeatedly received packets and

2 shows an information transmission with already transmitting side partial packet formation.

The regularly battery-powered transmitter 11 is, for example, the control unit of a radio remote control for building services equipment, but in particular a consumption meter (smart meter), which emits sporadically or quasi-continuously current or accumulated measured values. Accordingly, the receiver 12 can be the switching device for a drive device on site, or a concentrator for temporarily storing and, if appropriate, processing measured values which, corresponding to source coding, can also be received by a plurality of different measuring points. Likewise, it may also be a radio link 13 from a transmitter-equipped concentrator to a receiver at an addressed measuring point, for example to transmit tariff change information. In practice, these are predominantly such bi-directional radio links between a concentrator and a number of measuring points.

The exemplary unidirectional outlined information transfer from the transmitter 11 to the receiver 12 is successively packet by packet over a radio link 13 in one of the freely available, but correspondingly highly stressed and heavily interfered with ISM bands. The repeatedly transmitted information is in each case coded, for example, as a binary sequence to a packet 14 .i of 64 bytes in length.

Even if only one bit of a received packet 14.i is disturbed, the information is lost. Each packet 14.i is therefore transmitted i = x times. Conventionally, for the communication to succeed, at least two consecutive of the x repetitions must have been received completely undisturbed, ie exactly matching. The resulting transmission time required is thus significant, since in each case successively received complete packets 14 must be compared with each other, and possibly discarded.

On the other hand, as already explained, the invention is based on the consideration that, as a rule, only part of the received packet 14 is actually disturbed, while the undisturbed part corresponds exactly to the transmitted bit sequence.

Therefore, in the variant according to the invention according to FIG. 1, each received packet 14. I in the receiver 12 is split up into subpackets 15 .j = y and buffered. For this purpose, packets 14. I are received in succession. Subpackages 15. J assigned to one another are compared with one another and are considered to be correct, namely received as undisturbed, if at least two of them agree with one another. If thus all received subpackets 15.j (j = 1... Y) are recognized as valid, they are lined up to the undisturbed receive packet 14 '.

The subpackets 15.j within the packets 14.i can basically be of different lengths, their length can also vary. In the drawing, to simplify matters Position equal length subpackets 15.j outlines, for example (typically at least) j = y = 4 subpackets 15.j each 16 bytes of a packet 14. i of 64 bytes in length. The probability of transmitting a long subpacket 15 undisturbed is generally lower than it is in the case of a shorter subpacket 15 because of the longer duration of interference possibilities. Shorter (but not too short) subpackets 15.j thus provide increased probabilities regarding the existence of undisturbed subpackets 15J.

In the realization example shown in FIG. 1, x = i = 4 successive identical-content, complete packets 14 .i are received and buffered in the receiver 12. Those of at least two subpackets 15.j associated with each other with respect to their respective position in the packet 14.i, having at least twice matching contents (here 15.1 in 14.1 and 14.3; 15.2 in 14.1 and 14.3; 15.3 in 14.2 and 14.4; 15.4 in 14.2 and 14.4) are received as undisturbed and are therefore lined up to fill the valid undisturbed reception packet 14 '. Thus, a packet 14 'is available at the receiver 12 faster than if it waits until at least one entire packet repetition matches.

Even faster is a valid since undisturbed receive packet 14 'available if, according to the development according to Fig.2 already transmitting side a subdivision of the respective package 14.i in subpackets 15.j (j = 1 ... y) takes place. In order to determine whether these received subpackets 15.j are valid in each case, their bit sequence is tested on the receiving side for correctness. For this, each of these subpackets 15.i has an error detection code 16 already added at the transmitting end in a manner known as such, such as a CRC checksum.

Again, for the exemplary sketch, the subdivision in all packets 14.i to subpackets 15.j is assumed to be the same without having to be the same; namely again as a subdivision of the 64-byte packet 14. i each in four equally long sub-packets 5.y = 4 to 16 bytes.

In the receiver 12, each received subpacket 15.j of a packet 14, which has just been received, can be rejected immediately if the error detection code 16 detects a transmission error in this subpacket 15.j; here in the packet 14.i = 1 in the subpackets 15.j = 1 and 15.3 and in the sequence packet 14.2 in the subpackets 15.2 and 15.4. The remaining gaps in the receive-side packet 14 'remaining on receipt of the first packet 14.1 are filled up from the next (or, if necessary, after-next, etc.) received packet 14.i, in which not exactly this missing partial packet 15.j is disturbed again ; which then already the undisturbed receive packet 14 'lined up is available. Instead of or in addition to error detection 16, it is also possible to request that, in comparison to the method according to FIG. 1, the successive received and mutually compared partial packets 15 .j coincide with more than two transmissions in order to classify them as valid. This leads to increased reliability of the received data.

Another advantageous combination of the two measures described is to disassemble the error detection code 6 according to Figure 2 recognized as useful subpackets 15J turn sank in sub-subpackets and to select in mutual comparison according to Figure 1. This improves the residual error probability.

It is also possible, according to a further development of the invention, to combine the two measures described with respect to FIG. 1 or FIG. 2 in the sense that not only the partial packets 15. 1 shown as being usable via error detection codes 16 according to FIG. but also the useless ones. From these, useful sub-packets can then be picked out by way of sub-decomposition in mutual comparison according to FIG.

When different error detection methods are combined in such a way, for example, a CRC checksum can be examined first. In order for subpackets 15.j recognized as defective to be subdivided into sub-subpackets, they can be examined and reconstructed by means of another checksum-for example, a 3-out-of-6 encoding.

In the repeated transmission of information in the form of a coded packet 14. i from a transmitter 11 via a potentially disturbed radio link 13 to a receiver 12, which is operated in particular at the concentrator of a detection system for consumption readings, so is the relatively fast undisturbed receive packet 14 'to Available, if according to the invention undisturbed received subpackets 15.j are strung to undisturbed receiving packet 14 '. For this purpose, successively received packets 14 .i can be buffered at the receiver 12 in order to compare sub-packets 15 .j associated therewith with each other and, in the case of a match to the undisturbed reception packet 14 ', to be filled in a gap-filling manner. Or the successive packets 14.i are each already the transmitting side to - provided with error detection codes 16 - sub-packages disassembled 15.j, which are lined up filling the undisturbed receiving packet 14 ', if they were received correctly.

Claims

1. A method for transmitting information in the form of a coded packet (14) which is repeatedly transmitted from a transmitter (1 1) via a radio link (13) to a receiver (12), in which then an undisturbed receiving packet (14 ') is available, characterized in that undisturbed subpackets (15J) of the packets (14. i) are strung to undisturbed receiving packet (14 ').

2. The method according to claim 1, characterized in that the multiply successively transmitted packets (14. i) the reception side in the subpackets (15.j) are decomposed.

3. The method according to claim 1, characterized in that the packets (14.i) transmitted several times in succession are decomposed on the transmission side into subpackets (15.j) equipped with error detection codes (16).

4. The method according to any one of the preceding claims, characterized in that the lengths of the subpackets (15.j) are static or dynamically variable.

5. The method according to claim 2, characterized in that the receiving side each other with respect to their arrangement in the packet (14.i) associated sub-packets (15.j) successively received packets (14.i) compared to each other and lined up in accordance with the receiving packet (14 ') become.

6. Method according to the preceding claim, characterized in that at least x = 3 successively received packets (14.x) are each decomposed into a sequence of at least y = 3 subpackets (15.y).

7. The method according to claim 2, characterized in that the subpackets (15.j) are in turn divided into sub-sub-packages to be compared with each other.

8. The method according to claim 3, characterized in that also successively the error detection code (16) according to correctly received subpackets (15.j) are compared with each other for compliance. Method according to Claim 3, characterized in that the received subpackets (15.y) are in turn subdivided into sub-sub-packets to be compared with one another.

A method according to claim 3, characterized in that the subpackets (15.j) are provided on the sending side with CRC checksums, with 3-out-of-6 encodings or with other or with a plurality of error detection codes (16).